Recombinant adeno-associated virus (AAV) vectors can safely deliver long-term (>2 years), high-titers (>200 g/ml) of antibody-like inhibitors to rhesus macaques. These titers are well above the IC90s of several broadly neutralizing HIV-1 antibodies. AAV vectors have an established safety record in humans, and risks associated with their use are modest, arguably lower than those associated with current antiviral regimens. If key hurdles can be surmounted, AAV-expressed transgenes could replace less-tolerated components of current combination therapies, or supplant these therapies entirely. An AAV-based approach would avoid toxicities, costs, and compliance concerns associated with current therapies. However some challenges remain. First, more experience with these vectors will be necessary to further increase confidence in their safety. Second, the relative in vivo efficacy of well-known HIV-1 neutralizing antibodies remains poorly described, especially in combination. Third, HIV-1 can escape antibodies passively administered to infected humans, although limitations of antibodies tested may have accelerated this escape. Finally, a minority of AAV-inoculated immune-sufficient macaques raise antibodies to and thereby clear expressed transgenes. The core purpose of these projects is to show that advances in AAV-based gene therapy now make possible stable, long-term suppression of HIV-1 replication. Our immediate goal is to stably suppress an ongoing SHIV-infection in rhesus macaques, using specific combinations of antiviral proteins delivered by AAV vectors. To do so, we will address directly the key challenges of safety, efficacy, viral escape, and immune clearance of expressed transgenes. We have assembled a team with outstanding, complementary experience in macaque models of infection (Ronald Desrosiers, Harvard Medical School), therapeutic use of AAV transgenes (Guangping Gao, U Mass Medical School), HIV-1 entry and its inhibition with neutralizing antibodies (Michael Farzan, Harvard Medical School), and phage-based development and improvement of human antibodies (MSM Protein Technologies, led by Tajib Mirzabekov). We first focus on the common goal of establishing of a system in which viral loads in SHIV-infected animals can be reproducibly suppressed. Aspects of this baseline system will then be optimized to increase transgene efficacy (Projects 1 and 3), prevent immune clearance (Projects 1 and 2), and limit viral escape (Projects 3 and 4). These studies will establish principles and protocols directly applicable to subsequent human clinical trials.